Re: Folk Theorem (Steve_Kilbane)
Fri, 29 Oct 1993 14:32:44 GMT

          From comp.compilers

Related articles
Re: Folk Theorem: Assemblers are superior to Compilers (Mohd Hanafiah Abdullah) (1993-10-26)
Folk Theorem: Assemblers are superior to Compilers (1993-10-28)
Re: Folk Theorem (1993-10-29)
| List of all articles for this month |

Newsgroups: comp.compilers
From: (Steve_Kilbane)
Keywords: assembler, performance
Organization: Compilers Central
References: 93-10-114 93-10-131
Date: Fri, 29 Oct 1993 14:32:44 GMT

Hmm. I hope this thread isn't going to turn into a sequence of "my
favorite assembler hack" examples. While interesting, that wouldn't
accomplish much.

I agree that such uninformed articles as quoted constitute a RISK (to
quote another list), but not a huge one. When asked "will you write this
in assembler?", the answer is likely to mention huge increases in time and
cost ("Need to hire some good assembler coders"), and will also point out
that the customer had better not want to either change their design spec,
or want it to run on another platform. Ever. Such a reasoned argument
would put most clients off that particular idea.

Does an assembler hacker produce better code than a compiler? Well, this
depends on the target, the compiler, the application, the programmer, the
time/money/space/speed constraints, and so on. What's better? Yes, for
some occasions, use of assembler can be better than a compiler. But in
some cases, use of a formula 1 racing car is better than your average Ford
runabout. As always, apply your intelligence to your requirements, and
you'll have your answer.


Post a followup to this message

Return to the comp.compilers page.
Search the comp.compilers archives again.